On Sat, 23 Apr 2011, Ashish Sharma wrote:
I had a quick question about my regression of panel data. I am
using the fixed effects model to analyse data for 15 regions
over the span of 7 years. I do not want to hold time constant
but I do want to account for any regional differences between
regions by holding cross-sections fixed.
I am new to gretl and am not sure how to do this. When prompted
by gretl, before running my regression I did not tick the "time
dummy" box, however I cannot find an option for holding
cross-sections fixed.
Choose /Model/Panel/Fixed or random effects... and ensure that the
"fixed effects" button is selected (but it's the default so it
should be selected automatically).
Allin Cottrell